1. Field of the Invention
The present invention relates to an inspecting method employed for inspecting a surface of a magnetic recording medium of a hard disk drive (hereinafter, to be described as the HDD), more particularly to a technique for improving a detection efficiency of defects in a surface analysis test (SAT).
2. Background of the Invention
Various inspections are carried out for each HDD before delivery. Especially, a surface analysis test (SAT) is carried out as a defect test for the servo information written on a surface of each magnetic recording medium. In a usual SAT, a servo information defect is detected according to a decision of whether or not predetermined test pattern data can be normally written in a center of each track, for example, whether or not the data is normally written by the number of predetermined times. Whether or not the head following to the center of each track (controlling the movement of the head to above the center of the target track) is normally done is decided according to a PES (Positioning Error Signal) obtained by reading a burst pattern.
For example, Published Unexamined Patent Application No. 9-180388 and U.S. Pat. No. 5,872,677 specification have disclosed four burst patterns and a technique for generating a PES by reading those burst patterns. That is, both A and B burst patterns are written in the positional information areas at the same width as that of the track symmetrically about the center of the track on the surface of a subject medium. Both of C and D burst patterns are written in the positional information areas at the same width as that of the track symmetrically about the boundary between tracks on the surface of the medium. The burst patterns A and B written symmetrically about the center of each track generate an MPES (Main Positioning Error Signal) while the burst patterns C and D written symmetrically about each boundary between tracks generates a SPES (Slave Positioning Error Signal). Both MPES and SPES function complementarily with each other so as to be used as a signal representing positional deviation between tracks (PES). The PES referred to when in track following is an MPES. When the head moves from one track to another, however, the linear relations between both of the displacement and the MPES output cannot be kept as it is. Instead of the MPES, therefore, the SPES is used, since its sensitivity to the displacement at a boundary is the highest.
In the conventional writing test, a test pattern is written only in a write center of each track (the center of the track) repetitively and it is decided as a normal track (no error) when the test pattern can be written normally by the number of predetermined times.
However, while this method can detect MPES defects accurately around the center of each track, it cannot detect any SPES defects so accurately. This is because only one of C and D patterns exists around the center of each track, thereby the SPES value is not affected so much by the degradation of the burst pattern (C or D pattern) signal and even when the burst pattern could be hardly read. As a result, even when the C or D pattern includes a defect, the defect is not detected and the inspection is often continued as is.
On the other hand, in the case where any disturbance causes noise to be included in a signal related to the C or D burst that is not written actually in a target track, the magnitude relations between C and D are changed easily, thereby track numbers calculated according to MPES, SPES, and CYLID (cylinder number) are disordered. When such a disorder occurs, it often causes a serious error, for example, sudden head movement for compensation and/or execution of write operation in a wrong track.